nutrition-analyzer

Pass

Audited by Gen Agent Trust Hub on Jul 27, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection due to its processing of untrusted data from multiple dietary and health tracking sources without implementing sanitization or explicit boundary markers. An attacker could potentially embed malicious instructions within dietary log entries to influence the agent's behavior.
  • Ingestion points: The skill reads data from data-example/nutrition-logs/YYYY-MM/YYYY-MM-DD.json, data-example/profile.json, data-example/fitness-tracker.json, data-example/sleep-tracker.json, data-example/hypertension-tracker.json, and data-example/diabetes-tracker.json.
  • Boundary markers: No specific delimiters or instructions to disregard embedded commands are provided when the agent interpolates this external data into its context.
  • Capability inventory: The skill is granted Read, Grep, Glob, and Write tool permissions, which could be abused if an injection succeeds (e.g., unauthorized file modification).
  • Sanitization: There are no mentions of validation, filtering, or escaping for the ingested JSON content before it is processed.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 27, 2026, 08:21 PM
Security Audit — agent-trust-hub — nutrition-analyzer